GW2 Marketfansite

Guides/Starting out

All 36 elite specialisations, and how to choose without reading a meta list

Four per class, one per expansion, 250 hero points each. The full table, what an elite spec changes, and why the question "which is best" has a different answer every patch.

On this page
  1. The full table
  2. What an elite spec changes
  3. Why “which is best” has no stable answer
  4. How to choose, practically
  5. An observation about the design

Every class has four elite specialisations, one per expansion, which makes 36 in total. Each rewrites the class mechanic, adds a weapon and brings a new family of skills.

They cost 250 hero points apiece and unlock at 80 (where to get them fast).

The full table

Class Heart of Thorns Path of Fire End of Dragons Visions of Eternity
Guardian Dragonhunter Firebrand Willbender Luminary
Revenant Herald Renegade Vindicator Conduit
Warrior Berserker Spellbreaker Bladesworn Paragon
Engineer Scrapper Holosmith Mechanist Amalgam
Ranger Druid Soulbeast Untamed Galeshot
Thief Daredevil Deadeye Specter Antiquary
Elementalist Tempest Weaver Catalyst Evoker
Mesmer Chronomancer Mirage Virtuoso Troubadour
Necromancer Reaper Scourge Harbinger Ritualist

A specialisation needs its expansion. Owning End of Dragons does not unlock Firebrand.

What an elite spec changes

Three things, and the third is the one people underestimate.

A new weapon, often one the class could never hold.

A new skill family: shouts, wells, tomes, stances, whatever the theme wants.

The class mechanic itself. This is the real change. Necromancer Shroud becomes a melee form, or a sand field, or a health-draining trade (the four). Mesmer Virtuoso removes clones entirely (why that matters). You are not upgrading the class; you are replacing the part that made it that class.

Why “which is best” has no stable answer

Balance patches move roles around constantly. A specialisation that supplied Quickness for two years stops, another starts, and every tier list written before that patch becomes wrong without announcing itself.

So instead of a ranking that expires, the structural facts:

Group content asks for boons, not classes. A ten-player squad wants Quickness and Alacrity covered, plus healing (what those are). Which specialisations can do that changes; the requirement does not.

Most classes have at least one support option and one damage option. You are rarely forced off a class you like, only onto a different specialisation of it.

Open world does not care. Anything clears it. The question only sharpens in strikes and raids and high fractals.

How to choose, practically

Take the one that looks fun first. You will play it for weeks before the difference between it and the optimal choice matters to anybody.

Then take a second one for the other job. If your first is damage, make your second support, or the reverse. Two specialisations on one class cover most of what a group will ever ask of you, and switching between them costs nothing once unlocked.

Ignore the third and fourth until you have a reason. They will still be there.

Check the current answer when it matters. Before joining serious group content, look at what your class supplies this patch. Not before, and not from an article dated last year, including this one.

An observation about the design

Worth pausing on how unusual this system is.

Most games add power with expansions: the new thing is stronger, the old thing retires. Here each expansion added a lateral option. Core necromancer is still playable. Reaper did not replace it, and Harbinger did not replace Reaper.

The result is that a class you last played in 2015 has four versions of itself waiting, and the version you remember is one of them rather than an obsolete draft.

That also explains why the game asks 250 hero points instead of handing them over. The cost is not a barrier to power. It is a nudge to learn the core class before you replace its heart with something else.